© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 8589
    Prev
    Next
    INDEX
    Explanations

    XML tags and attributes

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/4-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.4.hook_resid_post
    Hook Layer
    4
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    343
    +0.12
    0.7%
    115
    +0.11
    0.7%
    204
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    78
    +0.12
    0.01
    9
    +0.11
    0.08
    397
    +0.11
    0.05
    Negative Logits
     Qatar
    -1.59
     Pakistan
    -1.51
    blogger
    -1.47
     lobby
    -1.44
     lncRNAs
    -1.43
     honey
    -1.39
    okers
    -1.38
     YouTube
    -1.36
     Wiki
    -1.35
     Pakistani
    -1.35
    POSITIVE LOGITS
    ·¸
    2.50
    µ
    2.48
    ´
    2.42
    ¤
    2.41
    ¬
    2.32
    č↵
    2.30
    ↵
    2.30
    ↵↵↵
    2.30
    <|outofrange|>
    2.30
     ↵
    2.30
    Activations Density 1.062%

    No Known Activations